By DAVE HOOK.
I’m not against UAVs. What I am against is fencing off parts of the sky from the flyers who were here first in the interest of UAV testing. I’m also not for rolling the dice with the lives of fellow aviators and their passengers.
A central challenge for integrating UAVs into the National Airspace System (NAS) safely is their current inability to see/sense-and-avoid. However, I look for solutions to situations given the limitations of existing science and technology. I have a solution to propose given what exists today.
Much of the see-and-avoid concept required of us under 14 CFR 91.113 depends upon both recognition of the potential for the collision with another aircraft and our rate of closure. That means we have an idea of the distance and speed of another aircraft based upon our sense of scale: A Boeing 747 that fills a particular space in our windscreen is certainly further away than the Piper Cub filling that same space.
So when aircraft of a smaller scale fill that same space ahead of us and we don’t recognize that difference in scale, we don’t recognize that the smaller aircraft is closer — requiring a faster response to avoid a mid-air collision.
To date there has been no systematic training to educate civil aviators in the United States about UAVs. Up to now we really haven’t had to worry much about them. But that’s about to change. There are six sites that are to begin UAV testing later this year. So where’s that safety training? You can only mitigate the risks you know about.
I’m absolutely against the use of TFRs in support of UAV testing. If the same principles of risk mitigation fence us out of UAV testing airspace, then why shouldn’t those same principles result in UAV operators not launching their UAVs, too? What’s good for the goose is good for the gander. It’s national airspace, not corporate airspace.
Here’s a potential solution for sharing the friendly skies with our remotely piloting colleagues that I wish to share.
I’m going to use as a basis for this solution a base altitude of 400 feet AGL. That’s because there is a long established practice for remotely controlled aircraft to be flown below 400 feet in accordance with FAA AC 91-57, Model Aircraft Operating Standards. That means any remotely piloted aircraft flying above 400 feet must be a UAV.
According to FAA Policy 8900.227, Unmanned Aircraft Systems (UAS) Operational Approval, a UAV that flies under IFR must have an operating Mode C transponder (a Mode S transponder is preferred). Go only half a step further and require that — as a minimum — a UAV always have a transponder with both Modes A and C. This means that the pilotless aircraft could be seen by ATC’s secondary radar system. Regardless of the size of the UAV and its radar cross section, the squawking UAV will be seen on the ATC controller’s radar scope, if the UAV is within line of sight of the radar antenna.
Next, don’t make the direct line of communication between the UAS controller and ATC preferred; make it a requirement in the Certificate of Authority or Special Airworthiness Certificate. If ATC detects conditions that could lead to a possible mid-air collision, they shouldn’t have to waste time establishing contact with the UAS controller. We’re talking safety here. Why wouldn’t you stack the odds in your favor?
Finally, make sure ATC can see a primary return on their radar scope from 400 feet AGL and up. That is, make sure they can detect a manned aircraft that is not squawking its transponder — referred to by the FAA as a non-cooperating aircraft.
Being able to see those two targets on their scopes affords ATC the opportunity to issue avoidance vectors to the UAV operator — a workaround for the see-and-avoid requirement in the NAS. There are, however, limits to all of this being possible.
The foundation to this solution is that the UAV must be able to be detected by ATC. This detection is easy to predict using the Radar Horizon Equation. For example, using a radar antenna height of 100 feet AGL and the RC-to-UAV transition altitude of 400 feet AGL, the maximum detection range at 400 feet AGL is approximately 37 nautical miles from the antenna. (For an explanation of why this works, click here. If you just want to enter numbers and let the webpage do the math, try Foruno U.S.A.’s Radar Horizon Calculator.)
Beyond 37 nm the radar controller can’t track aircraft down to 400 feet AGL. Yes, there can be obstacles and obscurants that could make the maximum distance less than 37 nm. Each UAV range would have to be considered individually based on existing conditions and obstructions, just like they are supposed to now.
The other reason why 37 nm works well is that distance is less than the 60 nm maximum detection distance for the current ASR-11 Air Traffic Control Radar System to pick up the primary (non-transponder) radar return of a typical small general aviation aircraft. With its transponder off or in standby, an aircraft can still be “seen” by ATC from only the reflected radar energy.
This approach does place a small manpower burden on the FAA’s ATC resources at the supporting radar facility. But it’s also the FAA that’s moving the UAV agenda forward — at the direction of Congress — without a fair sharing of the see-and-avoid responsibility between the pilots in manned aircraft and the UAV operators on the ground.
This approach of flying UAVs in such a way that ATC can issue avoidance vectors to UAS operators in lieu of see-and-avoid capability appears a practical way forward using existing capabilities within their limitations.
I’m sure there are alternate solutions that I haven’t considered. In fact, I believe that none of us is as smart as all of us. I hope that if there are other and better ideas out there among my readers that you will comment below and contribute your solution.
What I hope we can avoid is injury or loss of life during the upcoming UAV testing period and beyond.
Cheers from The Alamo.
I look at this from two perspectives. First, I for one am not totally against UAV’s (more commonly known as “drones”) and am not afraid of being in airspace with Boeing or Northrup or some smaller players with small aircraft models such as Insitiu. They will play by the rules, have the correct level of technology to maintain separation, and train operators well. But, no one else will. The individual private owners won’t. The farmers won’t. The paparazzi won’t. The flying Starbuck’s billboards that will be over every highway intersection won’t. The backyard fool with more money than sense will be flying formation on final with us and we won’t be able to do anything about it because there will be tens of thousands, maybe millions of these things everywhere. There will be no meaningful enforcement and all will be “flying” using their cell phones. For every UAV operator who behaves responsibly, there will be ten who won’t, won’t need to, and worse, won’t care. It is possible that airplanes will be regulated out of huge blocks of airspace, but that also it may become too dangerous to fly in any airspace in a highly populated area. License? We don’t need no stinking license!
The second problem that I see here is that it is possible and maybe even probable that if big enough commercial interests, such as Amazon, wants to use their drones in public airspace they would “buy” and privatize it. Seriously! Remember when the phone spectrum was auctioned off? How well did that work for all of us? It is not inconceivable that the airspace over LA or Seattle is worth billions. Buy off enough people in Congress or state legislatures, they pass laws, and it becomes legal to exclude you or other companies from their airspace. Their private commercial airspace that will be perfect for drones for pizza and Starbucks. Just automate it densely enough and it becomes possible and profitable. Then you not only can’t fly through it, but have no say in how its used and becomes another private commercial entitlement. Wouldn’t it be great to know that, like oil pipeline rights of way and private toll roads, you had the rights over your house rented to a TV company on the day of your daughter’s backyard wedding? Maybe you could “rent” back what you had jus a few years before! That day is coming. The commercial mandate from Congress t implement UAV’s wasn’t just an accident along the road of Progress.
The answer is obvious, make the UAV’s file and conform to a IFR flight plan. That way ATC can handle the “see and avoid” for them and our airspace remains safe.
Contrary to “Will says”,….The answer is not obvious. And simply filing an IFR flight plan does NOT solve the issue at all. ATC never had, does not now have, and never will likely have that authority to obviate the related pilot collision avoidance responsibility, dating back to the Wright’s first flight, including through the ICAO convention of 1944 and all its Annexes, to the present day. In the US, Read FAR 91.113 (b) for the relevant reference.
Active monitoring with a safety pilot nearby without TFRs isn’t the answer either (have you ever flown eastward into a blazing sunrise, or westward into a blazing sunset?). They’ll never be seen, which is how numerous aircraft have already collided with them, another transport just recently collided with one.
The only adequate solution from a safety perspective, that is possible by both the laws of physics and economics, is to use RNP based 4D trajectory separation. The UAV vehicles (10s of thousands of vehicles already out there) can just be too small to see (a few ft. wingspan), too high (well above FL300), and may even be flying with a negative groundspeed (sequencing WPs backwards) in a 120 knot wind field.
Dave Hook makes some key observations, and relevant suggestions for some very limited interim UAV operations. But the use of radar based vectoring of UAVs isn’t the long term answer. Although I strongly agree with Dave that TFRs aren’t the answer either.
Anybody who understands the physics of the problem, knows that FAA AVS’s concept of “Detect-Sense-and Avoid” doesn’t work, and can NEVER work, ever, for the majority of UAVs. The vehicle with the longer relative velocity vector controls the collision dynamic, and the velocity distributions of UAVs are typically the slower vehicle. That’s why birds or bugs who successfully detect, adequately sense, and nonetheless try to avoid you, by heroic maneuvers including rolling inverted, pulling 12 g’s, and diving at terminal velocity still just splat on you prop, windshield, strut, gear, or radome.
Instead, the only short and long term solution that obeys both the laws of physics and economics is to used automated dynamic RNP based 4D trajectory vehicle-vehicle separation via data links. It can work in conjunction with ATS, or it can still work even without ATS at all. This means of trajectory separation is inevitable in the global airspace system, but FAA just hasn’t figured it out yet. Which also is why NextGen is already turning into a $40B plus failure, as PastGen, based on FAA’s intending to use a flawed 1950’s concept, applying ADS-B simply as a “Pseudo radar” to provide 1:1 hand carried radar vector separation.
With use of modern FMSs, data links, and RNP, we could be starting to integrate UAVs into the global airspace system this year, not waiting until 2015, and we could do it without TFRs, or politically rigged test sites.
Dave,
Great article – and not a bad idea. However, you highlight one misunderstanding that many of us have about these test sites – and about UAS research operations in general: TFRs.
Many of us assume that the FAA will issue a TFR whenever UAS flights occur at test sites – not true. In fact, little will change from what happens today. The site administrators will still receive a COA, with the advantage that their application is expedited. They will still operate under line-of-sight, and will still use observers to ensure separation from other aircraft.
Groups like the University of North Dakota have been safely testing UAS operations for several years – without incident. And, in UND’s case, along side a flight school with over 100,00 hours of training flights flown each year.
In fact, many of these test sites are focused on technology to separate UAS aircraft, including the GPAR-RMS program (http://www.uasresearch.com/gpar-rms/default.aspx) and the LD-CAP ADS-B sense and avoid project (http://www.uasresearch.com/ld-cap/Default.aspx).
The current UAS operational rules require observers – many of whom are pilots. They are solely focused on identifying and mitigating traffic hazards and have a direct line of communication to the UAS operator. While ATC can also help identify hazards, they must split time between many tasks – making dedicated observers the safest bet.
It’s important that GA support the growing UAS industry – because we may be one of the biggest beneficiaries. Turbine and military aircraft were once the pinnacle of cockpit technology – not anymore. GA’s glass panel and navigation systems are often the envy of larger aircraft pilots – who still deal with 20 year old technology. But this technology came through developments in other industries, trickling down to GA.
UAS has the potential to do the same. From high bandwidth data links that can supply us with up to date weather data and digital ATC instructions, to advanced auto flight systems that were pioneered and scaled for mass UAS applications – GA stands to gain a lot.
As pilots, we should advocate for and support the UAS industry – and look at it with a positive attitude. They’re not out to get us – and most cases, they’re looking out for us.
Aleks
For Aleks benefit, ….while a laudable university experiment, and likely a very helpful activity to train students, both the Ganged Phased Array Radar-Risk Mitigation System (GPAR-RMS) and the other approach to use sensors onboard UAVs to identify and avoid other aircraft [Airborne Sense and Avoid (ABSAA), at best can hope for very limited utility. They will never serve but a tiny fraction of the UAV world, and never likely be deployed usefully. Even if deployed in small numbers, in a limited restricted environment, that can at best hope for very short and excessively expensive operational life. See the explanation above, in the other comments for why. Only RNP based 4D trajectory separation will work globally, and economically, for UAVs from bird size, moving at a few fps, to hypersonic vehicles during launch and recovery operating to the edge of space. Any other approach at this point, trying to further the FAA’s flawed “Detect-Sense-and Avoid” strategy, is likely simply wasted motion, delay, and money. It will turn into the next wasteful “MLS”. Serious UAV entities would be wise not to invest in this modern “D-S and A” FAA version of “fools gold”. It will not work for the predominant UAV world due to violating basic and fundamental laws of both the physics of collision dynamics, and installation economics.
Aleks, regarding your comment that “…GA’s glass panel and navigation systems are often the envy of larger aircraft pilots – who still deal with 20 year old technology” is very misleading. It might be the envy of a few DC-9-30 pilots still out there flying somewhere in Africa, or some AN-24 pilots flying in Russia, but your comment sure isn’t true in general of any modern large production jet transport, now typically equipped with EFIS, FMS, RNP levels down to .1, GLS, FANS, Data Links, SatComs, Cat III autoland, some with AIII HUDs, very effective digital WXR, TCAS traffic displays, IRUs and ADIRUs, … capable of global routing and operating in the worst weather. Even the very latest bizjets can’t come close to having either that capability, or the well designed globally operable human interfaces these transport jets sport. Hence your assertion of GA now being in the avionics lead,… is a frequently repeated (within in the GA community) case of wishful thinking, often on the part of some small and business aircraft avionics vendors, or business aircraft OEMs, but it is nonetheless largely incorrect, if not even being a complete myth.